Morphologically defined non-invasive breast lesions can be organized into categories, from those with little malignant potential (proliferative disease without atypia, PDWA), to those that may be 'initiated' (atypical ductal hyperplasia, ADH), to 'transformed' lesions that are considered to be malignant (ductal carcinoma in situ, DCIS). This series of progressively more abnormal lesions may be examples of evolutionary stages in the development of breast cancer, with each morphological stage representing the impact of cumulative genetic changes, culminating in a transformed cell lineage capable of overcoming controls on growth and differentiation. Our study of these early breast lesions (PDWA, ADH, DCIS), based on genetic loci known to show allelic imbalance (LOH) in invasive breast cancers, has indicated that the majority of these early lesions are themselves clonal neoplasms, the majority of early breast lesions can be thought of as ductal adenomas, with an evolutionary relationship to the invasive and that they commonly share many of the same LOH events found in invasive breast cancer in the same breast. Thus, the majority of early breast lesions can be thought of as ductal adenomas, with an evolutionary relationship to the invasive breast cancers that accompany them. That is, they represent true premalignant breast lesions. We therefore propose: (1) To seek common patterns in the emergence of LOH events as lesions evolve toward higher evolutionary stages, and also to seek patterns which may indicate the risk of a particular lesion's progressing to invasive cancer; (2) To confirm the association of particular LOH patterns in early lesions with progression to invasive disease, in order to guide clinical patient management decisions; and (3) To define in non-familial breast cancer specimens the role of tumor suppressor genes already identified in familial breast cancer, and also to study the role of these genes in the new families collected by our Familial Breast Cancer Registry (Resource B). The overall goal of this project is to identify genetic changes in premalignant breast cells which herald the progression to true malignancy and subsequent invasion, in order to more accurately diagnose and ultimately treat those premalignant lesions with a significant probability of becoming malignant. Measuring these genetic changes could identify patients who would benefit from chemo-prevention therapy. At the same time, we will learn more about the genetic loci which may be involved in the critical early stages of breast cancer evolution.